Invisible engineers and the history of computing

Dr Roger Johnson was a member of BCS Council for over 40 years and has seen the society go through times good and bad. As well as having accumulated a huge number of friends and contacts during his work with BCS, Roger has also gained a great deal of pleasure from some of BCS’ notable achievements over the years, as he explains.

What are some of your highlights?

Being president, something that you do for a year — apart from Maurice Wilkes — is a unique chance that comes along rarely. I was also president of CEPIS for two years, during which time we got the funding from the European Commission which really launched the European Computer Driving Licence. Dudley Dolan FBCS and I were largely responsible for the development of the business model, which proved to be phenomenally successful.

We were a pair of academics who had this idea — I think even we didn't understand the implications at the time — of making it a franchise. I remember sitting in a CEPIS meeting where we made a very optimistic target of getting a million licences, and I remember going to a celebratory dinner, probably in Brussels, where the then commissioner celebrated the millionth driving licence. And of course, it's gone into many, many millions now.

I also spent years representing BCS on IFIP and working closely with and between other member societies around the world, a major opportunity really to both contribute to, and spread, the BCS message of professionalism. Some of the societies already had it, but many had come out of a mathematical, scientific background and introducing these ideas to them was an interesting opportunity.

We know you well from the Computer Conservation Society. You were there from the start, weren’t you?

I was on the first committee of the Computer Conservation Society (CCS) and I had two roles; I was a founding member and then joined the committee. I was the technical vice-president.

The society started with Doron Swade, then curator of Computing at the Science Museum, and the CCS’ immediate past chair. Doron had the idea for a conservation society because he had old machines in the museum that he wanted to get running, so he came to BCS and was referred to the then technical director, Tony Sale. He was immediately enthusiastic about getting involved. Tony became secretary, Doron was the curator, and Ewart Willey was the first chairman of the group.

Maurice Wilkes was keen and was also, I think, one of the early speakers. He said, ‘I don't think this is going to last, we're going to run out of things to talk about’. Well, he continued to come until shortly before he died; he was a little hard of hearing, so he'd sit on the front row and I'd be wandering around as programme secretary and he would say ‘I told you it would never last’. He always remembered telling me that — he was a great guy.

I've also been secretary of the Conservation Society, and I've served one term of three years as chairman. I find it fascinating. I was intrigued from the outset and got involved with things like the Bombe rebuild with John Harper. I'm currently secretary of the Bombe Rebuild Trust at the National Museum of Computing at Bletchley.

I found as I was getting older that the technology was moving faster and faster, and if you're a computer scientist trying to give lectures, you really have to run to keep up. I found history went more slowly, so I got increasingly interested in the history of computing.

Tell us about what you call the 'invisible engineers'.

To me, the history of computing is important because — and this is one of the things that I stressed in my presidential year — the computing profession are the ‘invisible engineers’. I discovered an old manuscript of a talk, given during the year I was president, about the fact that we are invisible engineers because when we're successful, nobody knows we're there. People turn their terminals on, they go to their bank cash dispenser (or whatever) and it works.

This struck once when I had visited Vienna. Sitting in the departure lounge, there was the usual computer monitor on the desk — in those days producing cardboard boarding passes — and I could see out of the window the A320 (which was one of the first fly-by-wire commercial aircraft) that was going to bring us home to the UK. And I was thinking about how I had just checked in on a computer; how they are used for the food and drink going on to the plane; the baggage; air traffic control; the buying of tickets… and I realised that people look at the aircraft and think of aeronautical engineers. But not the computer engineers.

One of the things that isn't on my CV is that I was the very first person to buy a ticket over the internet from British Airways. I still have a certificate from them. They sent a stretch limo on a snowy morning in January to take my wife and me to Gatwick Airport and took our photograph.

Strangely, I found echoes of this during COVID. I mean, COVID was a disaster and a tragedy for so many people. And yet it was actually, in many respects, a huge triumph for IT. I'd worked in the university in a very small way, with some of the people who were involved with developing software to look at things like protein structures. The scientific work, such as the teams that were facilitating the vaccines, was hugely facilitated by IT. Then when it came to booking vaccinations, we had websites that popped up and said where the nearest vaccination venue was, complete with a list of dates and times to choose from. We went along at that date and time and lo and behold, the vaccines, by and large, were there waiting for us.

As a profession, we got remarkably little credit for what was a huge success. But we are noticed when things go wrong. The first time I wrote a letter as president of BCS in 1992, it was to The Times in response to the report on the London ambulance dispatch system failure. They had installed an online system to dispatch ambulances in response to calls, but in order to save money, they went to a small software house that had only built small-scale systems. They installed the system and there was a cascading failure. The whole system collapsed and calls went unanswered. The report showed that this did cost people’s lives.

One of the things that troubles me, is that we've not been very successful at educating the public about how dependent their lives are on these invisible engineers when it comes to making decisions about public policy, data protection, big data, AI and so on.

It must give you a fair amount of satisfaction, being so involved in the industry for all those years.

Yes. After my first four years, during which I worked for a software house building software that went into commercial use, I became an academic — and there’s a sense there that you're slightly a bystander, an observer, you're removed from the day-to-day nitty gritty.

One of the big changes over the years has been that leadership in IT has passed from academia largely into the commercial sphere and the number of researchers that the big social media companies and the like have is massive — far, far larger than any university department. When I moved from the big software house into academia, the academics were in the process of surrendering control. Academics still do contribute, but most innovation is driven by technology companies rather than academia.

I’ve been hugely lucky. If you read the early books, people used to say that you could have an all embracing World Computer Congress, which I still think is useful. But in those days you could cover the whole of computing. You'd have people talking about hardware, people talking about software, people talking about applications, and so on. Well-informed academics could still understand, at a quite decent level, the whole field of computing — but over my working lifetime, everybody has become a specialist.

I met Maurice Wilkes, I met Tommy Flowers of Colossus and I met quite a lot of early pioneers. People who did make a difference — it was a privilege to know them.

What are your thoughts on the progress of AI?

Durin my career I have met both Donald Mickie and James Lighthill, who wrote the report that brought about one of the AI winters. I also had a research project with Alexander D'Agapeyeff, who was BCS president and founded a major software house with Barney Gibbons — at which I happened to work. Alex was a huge enthusiast for expert systems in the mid-80s. He saw where things were going, and what we are now faced with in AI is something that has huge potential for good, but which is also capable of huge damage.

I've messed around with ChatGPT for fun, and it is a remarkable technology, but I was brought up in a politically aware, politically conscious household. I have a real concern about the capability of AI to industrialise the production of propaganda. It's going to be near impossible, at least with today's technology, for people to sort out whether information is true or false.

For you

Be part of something bigger, join BCS, The Chartered Institute for IT.

Last year, I met with Brian Randell, a retired professor from Newcastle. He has been researching Percy Ludgate, a relatively unknown Irish early 20th-century builder of mechanical calculators. I asked Brian to talk to the Conservation Society because he had asked Chat GPT to tell him about this little-known Irish inventor and got pages of answers to his questions. When he went through and he marked up the things that were right and things that were manifestly untrue, what alarmed him was a great slab of grey in the middle.

One of the things Chat GPT is well known for doing is that, if you ask it to write an academic paper, it will include very plausible references [which are nonetheless fake]. There is a well-known early example where a Nobel laureate and another very distinguished economist were credited as authors of an economics paper in a highly-rated journal. It gave the title of the paper, the journal, the volume and the page number. The only problem was that the Nobel laureate and his colleague hadn't written it, and the paper didn't exist. It was only when somebody went and looked it up that they discovered that ChatGPT had done what it always does, which is perform a kind of averaging process: who would write papers about this topic? What's a typical title? What’s a good journal for it to be in?

In Brian Randell’s example, ChatGPT returned a quote about Percy Ludgate from an Irish regional newspaper, the Cork Examiner. That’s a real paper, but after many, many hours spent searching, Brian could find no newspaper library in the world with a copy of that quote. So, has Chat GPT found a reference that nobody else can find? Or has it looked for a plausible Irish newspaper at the right sort of time and it is a complete invention? Who knows? It's very concerning, and we have to address the problem of industrialisation of potentially fake material.

The only consolation that I can draw on is my experience of looking at Alan Turing and the Bombe and the Enigma machine; it’s a battery-operated machine that a soldier in the field could type a message on and send by Morse code. It essentially industrialised encipherment — and all of a sudden what we needed to beat that machine was another machine. With AI, I think we have to industrialise the sifting of this material. I have spent a lot of time on the work of Andrew and Kathleen Booth, pioneers in the field of natural language translation, which has given me faith that because the AI material is mechanically and — ultimately — algorithmically produced, we can have sufficient confidence in the power of IT to provide us with tools to fight back.

How do you feel about the honorary fellowship?

Well, I'm greatly honoured to have been awarded it, and as I come towards the end of my active life within BCS, I feel very honoured to have that as a ‘thank you’ from a society to which I have devoted literally thousands of hours of my time on a voluntary basis. And it's a real pleasure to contribute. It’s rewarding to feel that I may have been able to help move all sorts of different projects forward a little bit. But I've been standing on the shoulders of giants. And if I have managed to contribute anything lasting, then to now have others metaphorically standing on my shoulders would be a great honour. Just as long as they don't do it physically, because I'm not sure I can hold them up!